Abnormal gut microbiota may cause PD-1 inhibitor-related cardiotoxicity via suppressing regulatory T cells

Cao, Huili; Dai, Huwei; Li, Songshan; Afzal, Zeeshan; Wang, Xinyan; Wen, Zeyu; Xiao, Kaiyong; Zhao, Yajing; Li, Jin; Yang, Bin

doi:10.1038/s41598-025-05635-4

Download PDF

Article
Open access
Published: 01 July 2025

Abnormal gut microbiota may cause PD-1 inhibitor-related cardiotoxicity via suppressing regulatory T cells

Huili Cao¹^na1,
Huwei Dai²^na1,
Songshan Li²,
Zeeshan Afzal²,
Xinyan Wang²,
Zeyu Wen²,
Kaiyong Xiao²,
Yajing Zhao¹,
Jin Li³ &
…
Bin Yang¹

Scientific Reports volume 15, Article number: 20547 (2025) Cite this article

Subjects

Abstract

Treatment with PD-1 inhibitors may cause immune-related adverse events (irAE), of which PD-1 inhibitor-associated myocarditis is a rare and highly lethal irAE. However, the mechanism of PD-1 inhibitors-induced myocarditis in individuals with tumor is still unclear. Regulatory T cells (Treg) can directly inhibit T cell proliferation and activation, and also produce inhibitory cytokines with potent immunosuppressive properties. Deletion or aberrant function of Treg usually leads to autoimmune diseases. But reports on the role of Treg in PD-1 inhibitor-associated myocarditis are still very limited, and its role in the pathogenesis of myocarditis needs to be further explored. In addition, alterations in the composition of the gut microbiota and its metabolites have been shown to be involved in the development of several cardiovascular and autoimmune diseases. Therefore, we designed this experiment initially to investigate the role of gut microbiota in PD-1 inhibitor-associated myocarditis. Our studies showed that PD-1 inhibitors induced myocarditis and significant reduction of intracardiac Tregs in melanoma mice. It is highly likely that alterations in the composition of the gut microbiota due to PD-1 inhibitors played a key role in this process.

Introduction

Immune checkpoint inhibitors (ICIs) are monoclonal antibodies that are used to stimulate the immune system against cancer cells. However, therapy with ICIs may produce immune-related adverse effects (irAE), of which ICI-associated myocarditis is an uncommon and very deadly irAE¹. Programmed cell death 1 (PD-1) is one of the approaches to blocking immune checkpoint; it inhibits T cells’ activation by specifically binding to programmed death ligand 1 (PD-L1). PD-1/PD-L1 pathway is crucial in regulating cardiac autoimmune responses. Mice with a deficiency of PD-1 produce high titers of autoantibodies against troponin I (cTnI), eventually leading to the development of autoimmune dilated cardiomyopathy². In vivo studies have shown that, animals treated with PD-1 inhibitors develop T-lymphocyte-infiltrated myocarditis^3,4. However, the mechanism of PD-1 inhibitor-induced myocarditis remains unclear, which has resulted in a lack of specific treatments for PD-1 inhibitor-associated myocarditis, as well as a high mortality rate in patients. Therefore, it is necessary to establish an animal model of PD-1 inhibitor-associated myocarditis that can mimic the pathophysiological state of tumor patients.

Regulatory T cells (Tregs) are a specific type of immunosuppressive T cells characterized by the expression of FOXP3, CD25, and CD4. Tregs can directly inhibit T cell proliferation and activation via cell contact, as well as produce inhibitory cytokines with significant immunosuppressive properties, including interleukin 10 (IL-10) and transforming growth factor beta (TGF-β)⁵. Tregs deficiency or dysfunction often results in the development of autoimmune disorders and exacerbation of inflammation. Previous studies have emphasized the critical role of Tregs in protection against acute viral or autoimmune myocarditis. Cao et al. found that after the removal of Tregs, myocardial hypertrophy and myocardial necrosis were significantly increased in mice with coxsackievirus B3-induced myocarditis⁶. Tajiri et al. discovered that reducing Tregs in the heart resulted in increased cardiac inflammation in mice with experimental autoimmune myocarditis⁷. However, reports on Tregs in Immune checkpoint inhibitor-associated myocarditis are still very limited, and its role in myocarditis development needs to be further explored.

In recent years, it has been found that gut microbes are closely linked to the body’s immunological response. Various factors, such as medication, nutrition, and exercise, can alter the composition and function of an individual’s gut microbiota and cause secondary effects, irrespective of the contributing factors themselves^8,9,10,11. Alterations in the gut microbiota composition, known as dysbiosis, and its metabolites have been shown to be significantly associated with the development of several cardiovascular and autoimmune disorders^12,13,14,15. In addition, the gut microbiota affects responsiveness to cancer therapy and susceptibility to toxic side effects¹⁰.

Here we report a mouse model to observe that the PD-1 inhibitors lead to a reduction in Tregs in the heart and a downregulation of the expression of a number of genes related to the function of Tregs. The second innovation is to test whether there is a correlation between some significantly altered gut microbes and a reduction in Tregs and genes. To test this, we constructed a melanoma mouse model and intervened with PD-1 inhibitors to explore the possible mechanisms of PD-1 inhibitor-induced myocarditis. We evaluated important immune cells, cytokine levels, and the composition of the gut microbiome under different interventions to further understand the possible pathways by which PD-1 inhibitors affect the development of myocarditis.

Results

PD-1 inhibitors induced myocarditis

H&E staining and Immunohistochemical staining reveals that the PD-1 inhibitors induced significant myocarditis compared to the Con group, by myocardial fibrous necrosis, fibrous tissue hyperplasia and infiltrations of segmental inflammatory cells. Anti group showed a higher number of CD4⁺ T cells, CD8⁺ T cells and macrophage infiltration than the Con group (Fig. 1A). And the Anti group exhibited more extensive myocardial injury according to the myocarditis severity score, including an increased area of myocarditis and elevated scores of CD4⁺ T cells, CD8⁺ T cells, and CD68⁺ macrophage infiltration (Fig. 1B). Additionally, the circulating CK-MB and cTn-I levels in Anti group were significantly elevated compared to Con group (Fig. 1C and D). In addition, the body weight and tumor volume of all mice were monitored throughout the experiment. Compared with the Con group, the tumor volume and body weight of mice in the Anti group showed a decreasing trend from day 8 and day 10, respectively (Fig. 1E and F).

PD-1 inhibition caused cardiac immune imbalance

Dysregulated immune homeostasis is a key factor in the development of myocarditis. Flow cytometry showed an increased number of CD4⁺ T cells and CD8⁺ T cells in the hearts of Anti group (Fig. 2A and B), however, the proportion of CD25⁺FOXP3⁺ Tregs was significantly decreased (Fig. 2C). Though the proportion of Th17 in the hearts of the Anti group was not significantly higher than in the Con group, the ratio of Th17 to Treg cells was significantly increased (Fig. 2D and E).

PD-1 inhibition down-regulated the expression of Tregs-related genes

Compared with the Con group, inhibition of PD-1 led to transcriptional downregulation of CD25, FOXP3, IL-10, and TGF-β in the hearts of mice, which are closely related to Tregs (Fig. 3A and D). Moreover, inhibition of PD-1 significantly reduced circulating concentrations of IL-10 and IL-17 (Fig. 3E and F).

PD-1 inhibition reprogrammed gut microbiota composition

To investigate the alterations in gut microbiota caused by PD-1 inhibition, we performed 16S rRNA gene sequencing of fecal samples from the Anti group and the Con group. A total of 203,844 high-quality reads were obtained. Alpha diversity, including Chao1, Shannon and Simpson indexes, was comparable between two groups (Fig. 4A and C). The results of PCoA analysis showed that the composition of gut microbiota was different between two groups (P < 0.05 in PERMANOVA results for both bray_curtis metrics and abund_jaccard metrics) (Fig. 4D and E). The PLS-DA showed that the bacterial communities of the two groups clustered separately (Fig. 4F).

A Venn diagram showed that 1107 and 841 OTUs were unique to the Con and Anti groups, respectively (Fig. 5D). The relative abundance of the gut microbiota in these two groups is represented in the histograms of phylum, family and genus, respectively (Fig. 4G and I). A total of 16 phyla were identified at the phylum level, among which Bacteroidota, Firmicutes_A, and Firmicutes_D dominated the intestinal microbiota of both the Con group and the Anti group, and the treatment with PD-1 inhibitors altered the relative abundance of these bacteria. At the family level, Lactobacillaceae, Rikenellaceae, Muribaculaceae, and Lachnospiraceae were the main bacterial families of the control group mice. PD-1 inhibition reduced the relative abundance of the three bacterial families except Rikenellaceae. At the genus level, we showed 15 of the most abundant microbiota, including Alistipes_A, Lactobacillus, Limosilactobacillus, etc. PD-1 inhibitor treatment also had different effects on the relative abundance of these microbiota.

To further identify gut microbiota that differed significantly between the Con and Anti groups, we performed a LEfSe analysis and set the effect size cut-off for LDA scores at 2.0. Antigroup mice showed a higher abundance of Rikenellaceae, Christensenellaceae, Ruminococcaceae, Gammaproteobacteria, Oscillospira, Flexispira, Ruminococcus and Streptococcus (Figs. 5B).

Gut microbes associated with myocarditis phenotypes

We analyzed the relationship between identified significantly enriched gut microbiota and myocarditis parameters in Anti group using Spearman correlation analysis. Unexpectedly, except for Gammaproteobacteria, almost all gut microbes enriched in Anti group mice showed a strong correlation with multiple myocarditis parameters. Notably, Flexispira enriched in mice treated with anti-PD1 antibody was significantly correlated with all cardiac injury factors (CD4⁺ T-lymphocytes, CD8⁺ T-lymphocytes, Th17 cells, CK-MB, and cTn-I) and was also negatively correlated with cardioprotective factors (serum IL-10, CD25 mRNA, FOXP3 mRNA, IL-10 mRNA, FOXP3 mRNA, IL-10 mRNA, and TGF-β mRNA). Ruminococcaceae was positively associated with all cardiac injury factors except Th17 cells and negatively associated with all cardioprotective factors, including Tregs (Fig. 5C).

Fecal microbiota removal attenuates PD-1 Inhibitor-induced myocarditis

To further clarify the relationship between the gut microbiota and PD-1 inhibitor-associated myocarditis, we performed fecal microbial removal experiments. Compared with the Anti group, the area of myocardial injury and the proportion or score of inflammatory cells, such as CD4⁺ T, CD8⁺ T, and macrophages, were significantly reduced in the hearts of mice in the FMR group (Figs. 1A and B and 2A and B). Moreover, serum CK-MB and cTn-I concentrations were significantly reduced in the FMR group of mice (Fig. 2C and D). In addition, the trend of tumor volume changes in the FMR group was similar to that in the Anti group, and the body weights of mice in the FMR group showed a significant increase compared to those in the Anti group from day 10 (Fig. 1E and F).

Subsequently, we investigated the mechanisms underlying the attenuation of myocarditis by FMR. First, flow cytometry analysis showed that the ratio of CD25 + FOXP3 + Tregs was increased and the ratio of Th17/Treg was decreased in the myocardium of the FMR group relative to the Anti group, but the ratio of Th17 cells did not show significant changes (Fig. 2C and E). Secondly, the expression of Tregs-related genes, including CD25, FOXP3, IL-10 and TGF-β, were significantly up-regulated in the FMR group compared with the Anti group (Fig. 3A and D). In addition, the concentration of circulating IL-10 in the FMR group also appeared significantly elevated relative to the Anti group, but IL-17 did not change significantly (Fig. 3D and E).

FMR removed key microbiota associated with PD-1 Inhibitor-induced myocarditis

We analyzed the gut microbiota of FMR group by 16S rRNA gene sequencing to clarify the characteristics of its gut microbiota. As can be seen from the relative abundance histogram and Venn diagram of the microbiota in the FMR group, almost all the gut microbiota of the mice were in a state of extreme inhibition after antibiotic treatment. The gut environment of mice in the FMR group was composed of only a few types of microbiota, such as Proteobacteria, Enterobacteriaceae_A and Proteus, which constituted the main part of phylum, family and genus, respectively (Fig. 4G and I). And only five OTUs in the FMR group were the same as those in the Anti group (Fig. 5A). Moreover, LEfSe analysis showed that compared to the Anti group, The relative abundances of Rikenellaceae, Christensenellaceae, Ruminococcaceae, Oscillospira, Flexispira, Ruminococcus and Streptococcus in FMR group were significantly decreased, but that of Enterobacteriaceae_A, Gammaproteobacteria and Proteobacteria were significantly increased (Fig. 5D). Furthermore, the result of Spearman correlation analysis confirmed that Enterobacteriaceae_A, Gammaproteobacteria and Proteobacteria did not correlate with the parameters of PD-1 inhibitor-induced myocarditis. Thus, FMR cleared the key microbiota associated with PD-1 inhibitor-induced myocarditis.

Discussion

The gut microbiota has been shown to have a large impact on a wide range of immune and cardiovascular diseases. This study identified the critical role of gut microbiota in PD-1 inhibitor-associated myocarditis and firstly explored the association between gut microbiota and cardiac Tregs. In this study, the number of CD25⁺FOXP3⁺Treg and the transcription of genes such as IL-10 and TGF-β were significantly reduced in the hearts of mice treated with PD-1 inhibitors. And PD-1 inhibition also significantly altered the gut microbiota, which showed a strong correlation with the myocarditis phenotype. In addition, the aforementioned changes in Treg and related genes induced by PD-1 inhibitors could be suppressed by depletion of the gut microbiota. Therefore, it is reasonable to assume that the gut microbiota contributes, at least in part, to the deterioration of PD-1 inhibitor-associated myocarditis.

Our study demonstrated that PD-1 inhibitors induced myocarditis in mice with melanoma. Antigroup mice showed a significant increase in the number of inflammatory cells such as CD4⁺ T cells, CD8⁺ T cells, and macrophages in the myocardium, as well as a significant increase in serum CK-MB and cTn-I concentrations, which is consistent with the pathologic features of ICI-associated myocarditis described in clinical case reports^16,17,18. Past studies have reported that mice treated with PD-1 knockout or anti-PD-1 exhibit dilated cardiomyopathy with heart failure or only mild myocarditis. We therefore hypothesized that the presence of tumors may be a persistent immunostimulatory factor causing PD-1 inhibitor-associated myocarditis^19,20. In addition, the ratio of Th17 to Treg and the concentration of serum IL-17 were elevated in the hearts of mice treated with PD-1 inhibitors, which is consistent with the report by Tamas G et al.²⁰. In studies of viral myocarditis and autoimmune myocarditis, disruption of the balance between Th17 and Treg is often used to assess the direction of progression of myocarditis²¹, and this does seem to apply to myocarditis caused by PD-1 inhibitors. Based on our findings, it appears that the significant decrease in Tregs and the relative increase in Th17 cells may be the key factors leading to immune homeostasis disruption.

Subsequently, we analyzed genes in the myocardium that are closely related to the immunosuppressive function of Tregs. The results showed that the transcript levels of the CD25 gene, FOXP3 gene, IL-10 gene, and TGF-β gene were significantly downregulated. CD25 and FOXP3 are both specific markers for Tregs, and CD25 is closely associated with the suppressive activity of Tregs²², whereas FOXP3 is essential for the maturation and function of Tregs, and the lack of FOXP3 leads to the loss of Tregs function²³. TGF-β signaling by Tregs can limit the function of self-reactive T cells^24,25,26. Similar to TGF-β, IL-10 is important for the inhibitory capacity of Tregs, especially in regulating inflammatory responses triggered by pathogens or foreign particles²⁷. Thus, downregulation of the expression of these genes after treatment with PD-1 inhibitors negatively affects the differentiation, maturation, and anti-inflammatory functioning of Tregs, which may contribute to the disruption of local immune homeostasis in the myocardium and the development of myocarditis. In addition, we found that serum IL-10 was also significantly reduced in mice of Anti group, suggesting that PD-1 inhibitors systemically impaired the immunosuppressive effects of IL-10 on the organism, which may also be involved in the occurrence of other irAEs.

Changes in the composition of gut microbiota can have different effects on the body. Alterations in the gut microbiota can cause disruption of the barrier function of the gut, resulting in bacterial and endotoxin translocation^28,29. Myosin-peptide mimics derived from gut bacteria can promote autoimmune response against the heart^30,31. Furthermore, the gut microbiota also influences the host immune system by releasing metabolites with immunomodulatory functions. For example, short-chain fatty acids (SCFAs) exert anti-inflammatory effects by activating Tregs, which attenuates cardiac hypertrophy and fibrosis³². Trimethylamine N-oxide (TMAO) exacerbates cardiac fibrosis through activation of NLRP3 inflammasome³³. So, we used 16S rRNA gene sequencing analysis to compare differences in the composition of the gut microbiota between normal mice and anti-PD-1 mice. The α-diversity analysis of the gut microbiota indicated that the diversity was similar in both groups, but the application of PD-1 inhibitors altered the composition of the gut microbiota. Melanoma mice treated with PD-1 inhibitors showed an increased relative abundance of various microbiota at different levels, including Gammaproteobacteria, Rikenellaceae, Christensenellaceae, Ruminococcaceae, Oscillospira, Flexispira, Ruminococcus, and Streptococcus.

Spearman correlation analysis (Fig. 5B) revealed that, except for Gammaproteobacteria, the enriched gut microbiota were significantly positively correlated with myocardial inflammatory injury factors and significantly negatively correlated with protective factors at different levels, especially Ruminococcus and Flexispira. In previous studies, it has also been demonstrated that these gut microbes are closely associated with immune disorders and inflammatory diseases. For instance, reductions in Rikenellaceae and Oscillospira contribute to the alleviation of inflammatory bowel disease (IBD)^34,35. The enrichment of Christensenellaceae was associated with sepsis and acute intestinal injury³⁶. Ruminococcus promotes Crohn’s disease through the production of inflammatory polysaccharides³⁷, and Streptococcus contributes to rheumatic heart disease and myocarditis through molecular modeling^38,39. Flexispira is a pathogenic bacterium closely related to Helicobacter, which causes chronic bacteremia in patients with primary immune disorders X chain of apheresis^40,41. In addition, the enrichment of Flexispira may have promoted the differentiation of intestinal lamina propria lymphocytes toward Th17 cells⁴², which is consistent with the results of our correlation analysis. Although no definitive association has been established between Flexispira and myocarditis, based on the current evidence, further experiments to validate its close association with cardiac Th17 seems necessary. Moreover, in previous studies Ruminococcaceae were considered to be strongly associated with IBD and often coincided with increased disease activity^43,44. Therefore, in conjunction with the results of Spearman’s correlation analysis in this study, we hypothesized that Ruminococcaceae was most likely involved in disrupting the protective effects of Tregs, which promoted the progression of myocarditis in mice. And this seems to have been confirmed in another study, in which melanoma mice treated with BMS-1 (An inhibitor against PD-1 and PD-L1) also showed significant abundance enrichment of gut Ruminococcaceae, and altered gut microbes amplified the cardiotoxicity associated with BMS-1¹⁵.

Finally, to further determine the role of the enriched microbiota in PD-1 inhibitor-related myocarditis, we simultaneously conducted antibiotic intervention in the Anti group mice. The FMR experiment results indicated that Abx treatment effectively inhibited the reproduction of the gut microbiota in mice and significantly alleviated the cardiotoxicity caused by PD-1 inhibitors. In the hearts of FMR group mice, both the degree of myocardial injury and the proportion of inflammatory cells infiltrating the heart significantly decreased, accompanied by a significant upregulation of CD25⁺ FOXP3⁺ Tregs, serum IL-10 and the transcriptions of Treg-related genes. Moreover, it is worth emphasizing that although the ratio of Th17/Tregs in the hearts of FMR group mice was significantly lower than that of the Anti group mice, the concentration of serum IL-17 and the proportion of Th17 cells did not change significantly. Therefore, we believe that the impact of fecal microbiota removal on cardiac immune balance is more likely achieved through the regulation of Tregs. Moreover, the tumor volume of mice in the FMR group showed a decreasing trend during PD-1 inhibitor treatment, suggesting that antibiotics did not diminish the antitumor effect of PD-1 inhibitors, and that intervention in the gut microbiota may be a potentially effective method for preventing or treating PD-1 inhibitor-associated myocarditis in oncology patients. Additionally, 16S rRNA gene sequencing analysis showed that the total number and composition of gut microbiota in FMR group mice were completely different from those in the Anti group mice. Compared with the Anti group, we found a significant decrease in the relative abundance of all microbiota (Rikenellaceae, Christensenellaceae, Ruminococcaceae, Oscillospira, Flexispira, Ruminococcus and Streptococcus) closely associated with the myocarditis phenotype and an increase in the relative abundance of Enterobacteriaceae_A, Gammaproteobacteria, and Proteobacteria in the mice of FMR group (Fig. 5D). However, Spearman analysis results suggested that Enterobacteriaceae_A, Gammaproteobacteria and Proteobacteria were not correlated with the myocarditis phenotype (Fig. 5E). Therefore, the results of the FMR experiment further confirmed our speculation that the removal of key enriched microbiota can attenuate PD-1 inhibitor-related myocarditis, and this may be achieved by significantly increasing the number and immunosuppressive function of CD25⁺ FOXP3⁺ Tregs. However, more in-depth studies and larger sample sizes are needed to determine the exact mechanism of this effect.

This study suggests that gut microbiota may amplify PD-1 inhibitor-associated myocarditis by inhibiting the anti-inflammatory function of cardiac Tregs. However, our study still has some limitations. First, because of the limited number of experimental animals in this study, we were sometimes unable to collect enough mouse tissue samples for the full range of analysis, and the resulting conclusions need to be further confirmed by studies with larger sample sizes. And further rigorous experimental modeling is needed to clarify these biometric analysis results from molecular signaling experiments. Furthermore, in the FMR experiments, there is still a lack of a control group to clarify the effects of Abx treatment on myocarditis. In addition, because 16S rRNA gene sequencing methods make it difficult to identify most microbiota at the species level, further macrogenomic sequencing is needed to identify key microbial species and to clarify the effects of these microorganisms through microbial gut transplantation experiments.

Conclusion

In summary, our study results suggest that PD-1 inhibitors modify the composition of the gut microbiota in mice. The gut microbiota promotes PD-1 inhibitor-associated myocarditis by altering the number and function of Tregs. The gut microbiota may play an important role in potential therapeutic strategies to mitigate immune checkpoint inhibitor-associated cardiotoxicity.

Materials and methods

Drugs and reagents

For cell culture, we used fetal bovine serum, high-sugar Duchenne’s modified eagle’s medium, penicillin and streptomycin mixtures. The mouse IgG and RMP1-14 were used to establish the mouse model. We used anti-CD4 antibody, anti-CD8 antibody, and anti-CD68 antibody for immunohistochemical staining. The ELISA kits, including the Mouse cTn-I ELISA Kit, the Mouse CK-MB ELISA Kit, the Mouse IL-17 ELISA Kit and the Mouse IL-10 ELISA Kit, were used to detect the serum concentration of cTn-I, CK-MB IL-17 and IL-10. The anti-CD45 antibody, anti-CD4 antibody, anti-CD8 antibody, anti-CD25 antibody, anti-FOXP3 antibody, and anti-IL-17A antibody were used for flow cytometry analysis. A cocktail of vancomycin, neomycin sulfate, and primaxin was used to clear the gut microbiota in the FMR group. All the drugs and reagents that were used in the experiment are listed in Table 1.

Table 1 Highlights the details of the drugs and reagents that were used in the experiment.

Full size table

Cell culture

The B16-F10 mouse melanoma cells were procured from the American Type Culture Collection (USA) and cultured in high-sugar Duchenne’s modified eagle’s medium (4.5 g/L) containing 10% fetal bovine serum and 100 U/ml penicillin and streptomycin. All cells were cultured at 37 °C in a humidified 5% CO₂ atmosphere.

Animal model and treatment

BALB/c male mice, aged between 6 and 8 weeks, were purchased from Chengdu Dashuo Experimental Animal Co., Ltd. and bred in the laboratory of the Department of Cardiology, Second Hospital of Shanxi Medical University. All animal experimental protocols were approved by the Ethics Committee of the Second Hospital of Shanxi Medical University (DW2022081) and all methods were carried out in accordance with relevant guidelines and regulations. All methods are reported in accordance with ARRIVE guidelines.

B16-F10 cells (1 × 10⁶) were first injected into the right axilla of BALB/c mice to establish the model. When the tumor size reached approximately 100 mm³, the mice (n = 18) were randomly divided into three groups: the Anti-PD1 (Anti) group (n = 6), the Isotype control (Con) group (n = 6) and fecal microbiota removal (FMR) group (n = 6).

Starting from day 0, the Anti group and FMR group received intraperitoneal injections of 250 µg anti-PD1 antibody (RMP1-14): a single dose of injection given every two days apart for a total of 7 doses to induce myocarditis, while a single dose of mouse IgG injection was given every two days apart for a total of 7 doses intraperitoneally into the Con group.

Fecal microbiota removal (FMR): From Day 0, a cocktail of broad-spectrum antibiotics (Abx) containing 0.5 g/L Vancomycin, 0.5 g/L Neomycin Sulfate, and 0.5 g/L Primaxin was added to the drinking water of mice in the FMR group to clear the gut microbiota. On day 14, all the mice were euthanized, and tissue samples were obtained for examination. Fresh fecal samples were collected in sterile tubes and immediately frozen in liquid nitrogen, then stored at −80 °C for subsequent DNA extraction and 16S rRNA gene amplification.

After anesthetizing mice with inhaled isoflurane, blood was drawn from the heart for a subsequent enzyme-linked immunosorbent assay (ELISA). The mice were then euthanized by a rapid blood draw through the heart under deep anesthesia. The hearts were removed and fully lavaged with sterile phosphate buffered solution (PBS) at 4 °C to clear residual blood for subsequent histopathological analysis, flow cytometry analysis, and quantitative real-time polymerase chain reaction (qPCR). A detailed stepwise process for the establishment of animal model is shown in Fig. 6.

Histopathological analysis

H&E staining

The necropsy was conducted on all the mice. The cardiac tissues were soaked in 4% paraformaldehyde at room temperature for 48 h, fixed in paraffin, sliced into 3 μm sections, and stained with hematoxylin and eosin (H&E) for microscopic examination. Then, images of 100× cardiac tissue were captured by a Pannoramic 250 digital section scanner (3DHISTECH), and the total area of myocardial lesions was measured using Image Pro Plus 6.0.

Immunohistochemical staining

The cardiac tissues were fixed in 4% paraformaldehyde at room temperature for 48 h, embedded in paraffin, and sliced into 4 μm sections. These sections were then deparaffinized, rehydrated, and treated with the sodium citrate-hydrochloric acid buffer solution (pH 6.0) in a microwave oven. After that, the sections were blocked with 10% goat serum (Bosterbio, China) at room temperature for 20 min, followed by incubation overnight at 4 °C with rabbit anti-mouse primary antibodies to CD4, CD8, and CD68 (1:100 dilution), respectively. After being rinsed with PBS, they were incubated with horseradish peroxidase (HRP)-labeled goat anti-rabbit secondary antibodies (1:100 dilution) at 37 °C in the dark for 30 min. Finally, after the tissues were developed at room temperature using the DAB kit (ZSGB-bio, China), hematoxylin was used to re-stain the nuclei of the cells for 3 min. Images were obtained under a digital trinocular camera microscope (McAudi, China) at a magnification of ×40, and positive results were quantified using Halo 101-WL-HALO-1 digital image analysis systems (Indica Labs, USA).

Serum biochemistry analysis

The blood removed from the mice was immediately centrifuged for 5 min (3000 rpm, 4 °C), and the serum was collected from the upper layer. Afterwards, ELISA was performed according to the kit instructions. Serum levels of CK-MB, cTn-I, IL-17 and IL-10 were calculated based on absorbance (OD) measured at 450 nm by an enzyme marker (Molecular Devices, China).

Flow cytometry

Immune cell isolation

The ventricular tissues were cut into 5 mm thick pieces and placed in a digestion solution consisting of 1.5 mg/ml Collagenase A (Sigma, USA), 0.4 mg/ml DNase I (Sigma, USA), 5% fetal bovine serum (ThermoFisher, USA), and 10 mM HEPES (ThermoFisher, USA) and shaken (250 rpm, 45 min) at 37 °C. After digestion, the cells were mixed with PBS and filtered through a 70 μm cell separator. Later, the remaining immune cells were resuspended in PBS for subsequent flow cytometry analysis.

Flow cytometry analysis

After resuspension of the cells in 100 µL PBS, the required antibodies were added and incubated for 30 min at 4 °C, protected from light. The antibodies and doses for each sample were as follows: Anti-CD45 antibody (1.25µL), Anti-CD3 antibody (2µL), Anti-CD4 antibody (2.5µL), Anti-CD8 antibody (2.5µL), Anti-CD25 antibody (2.5µL), Anti-IL-17 antibody (1.25µL). 1x true nuclear fixation concentrate (500µL) was then added and incubated for 50 min at room temperature before centrifugation (350 rpm, 5 min). After discarding the supernatant, the cells were washed twice and resuspended with 1x True Nuclear Perm. After discarding the supernatant, the resulting cells were washed twice and resuspended in 1x True Nuclear Perm. Anti-FOXP3 antibody (5 µL) was added and incubated at 4 °C for 30 min, washed with 1x True Nuclear Perm, and centrifuged (350 rpm, 5 min). The resulting cells were resuspended with 1x True Nuclear Perm and detected using a flow cytometer (Beckman, USA). In CytExpert, the results were analyzed.

Realtime fluorescence quantitative polymerase chain reaction

The heart’s total RNA was extracted using an animal total RNA isolation kit (FOREGENE, USA). cDNA was synthesized by the PrimeScript RT reagent kit (Takara Bio, China). qPCR was performed using 2×Real PCR EasyTM Mix-SYBR (Takara Bio, China), and different primer sets are listed in Table 2. Reactions were performed using the QuantStudio TM3 Real-Time PCR Detection System (ThermoFisher, USA). Analysis of the threshold cycle for each sample of the PCR process was performed using Thermo Scientific PikoReal (Thermo, USA). The relative expression of each specific gene was normalized to 2 ^△△CT (△CT = specific gene CT - β-actin CT; △△CT=△CT of experimental group specific gene -△CT of control group specific gene).

Table 2 Shows the different primer sequences for qPCR.

Full size table

16S rRNA gene sequence analysis

Fecal bacterial DNA was extracted using the E.Z.N.A.^® Stool DNA Kit (Omega Biotech, USA) according to the manufacturer’s protocol. Nanodrops quantified DNA, and 1.2% agarose gel electrophoresis checked the quality of DNA extraction. The highly variable V3-V4 region of the bacterial 16S rRNA gene was amplified by PCR using a set of primers (338 F 5′ACTCC TACGGGAGGCAGCAG-3′, 806R 5′GGACTACHVGGGT WTCTAAT-3′). The amplification products were then purified and recovered using Vazyme VAHTSTMTM DNA-clean beads. Following this, the recovered PCR amplification products were fluorescently quantified using the Quant-iT PicoGreen dsDNA Assay Kit (Thermofisher Scientific, China) with a microplate reader (BioTek FLx800, USA). According to Majorbio Bio-Pharm Technology Co., Ltd.‘s standard protocol, the purified amplicons were mixed in the appropriate proportions based on the fluorescence quantification results. Finally, sequencing libraries were prepared using the TruSeq Nano DNA LT Library Prep Kit (Illumina, USA), and DNA fragments were paired-end sequenced on the Illumina MiSeq platform (Illumina, USA).

The raw FASTQ files were quality-controlled, denoised, spliced, and de-chimerized by QIIME2 (2019.4, https://qiime2.org/) calling DADA2. The resulting sequences were subsequently clustered into amplicon sequence variants (ASVs) with 100% similarity. The taxonomic species were annotated using the GreenGenes database (Release 13.8, http://greengenes.secondgenome.com/)⁴⁵. The ggplot2 packages in R (Version 4.3.0, https://www.r-project.org/) were used to perform taxonomic compositional analysis. The analysis of alpha diversity indices, including Chao1, Simpson, and Shannon, beta diversity indices represented by principal coordinate analysis (PCoA) projections, and partial least squares-discriminate analysis (PLS-DA) were done in R script. Bacterial differences between groups were determined using linear discriminant analysis (LDA) and effect size (LEfSe) with the R ggtree and ggplot2 packages. Only LDA values > 2.0 with a p-value of 0.05 were considered to be significantly enriched.

Statistical analysis

Data analysis was performed using Graphpad Prism 9.3.0. The data were presented as mean ± SD. The One-way ANOVA test was used for statistical analysis. A P-value less than 0.05 was considered statistically significant.

Data availability

The datasets of 16sRNA sequences generated during the current study are available in the NCBI repository, [https://www.ncbi.nlm.nih.gov/sra/PRJNA1154955].

References

Hu, J. R. et al. Cardiovascular toxicities associated with immune checkpoint inhibitors. Cardiovasc. Res. 115, 854–868 (2019).
Article CAS Google Scholar
Okazaki, T. et al. Autoantibodies against cardiac troponin I are responsible for dilated cardiomyopathy in PD-1-deficient mice. Nat. Med. 9, 1477–1483 (2003).
Article CAS Google Scholar
Ji, C. et al. Myocarditis in Cynomolgus monkeys following treatment with immune checkpoint inhibitors. Clin. Cancer Res. 25, 4735–4748 (2019).
Article CAS Google Scholar
Michel, L. et al. Targeting early stages of cardiotoxicity from anti-PD1 immune checkpoint inhibitor therapy. Eur. Heart J. 43, 316–329 (2022).
Article CAS Google Scholar
Sakaguchi, S. Naturally arising Foxp3-expressing CD25 + CD4 + regulatory T cells in immunological tolerance to self and non-self. Nat. Immunol. 6, 345–352 (2005).
Article CAS Google Scholar
Cao, Y., Xu, W. & Xiong, S. Adoptive transfer of regulatory T cells protects against coxsackievirus B3-induced cardiac fibrosis. PLoS One. 8, e74955 (2013).
Article ADS CAS Google Scholar
Tajiri, K. et al. Endothelin receptor antagonist exacerbates autoimmune myocarditis in mice. Life Sci. 118, 288–296 (2014).
Article CAS Google Scholar
Cresci, G. A. & Bawden, E. Gut microbiome: what we do and don’t know. Nutr. Clin. Pract. 30, 734–746 (2015).
Article CAS Google Scholar
Hall, A. B., Tolonen, A. C. & Xavier, R. J. Human genetic variation and the gut Microbiome in disease. Nat. Rev. Genet. 18, 690–699 (2017).
Article CAS Google Scholar
Roy, S. & Trinchieri, G. Microbiota: a key orchestrator of cancer therapy. Nat. Rev. Cancer. 17, 271–285 (2017).
Article CAS Google Scholar
Hitchings, R. & Kelly, L. Predicting and Understanding the human microbiome’s impact on Pharmacology. Trends Pharmacol. Sci. 40, 495–505 (2019).
Article CAS Google Scholar
Cekanaviciute, E. et al. Gut bacteria from multiple sclerosis patients modulate human T cells and exacerbate symptoms in mouse models. Proc. Natl. Acad. Sci. U S A. 114, 10713–10718 (2017).
Article ADS CAS Google Scholar
Kasahara, K. & Rey, F. E. The emerging role of gut microbial metabolism on cardiovascular disease. Curr. Opin. Microbiol. 50, 64–70 (2019).
Article CAS Google Scholar
Wu, Q. et al. Gut microbiota can impact chronic murine lung allograft rejection. Am. J. Respir Cell. Mol. Biol. 60, 131–134 (2019).
Article CAS Google Scholar
Chen, Y. et al. Prevotellaceae produces butyrate to alleviate PD-1/PD-L1 inhibitor-related cardiotoxicity via PPARα-CYP4X1 axis in colonic macrophages. J. Exp. Clin. Cancer Res. 41, 1 (2022).
Article Google Scholar
Johnson, D. B. et al. Fulminant myocarditis with combination immune checkpoint Blockade. N Engl. J. Med. 375, 1749–1755 (2016).
Article Google Scholar
Norwood, T. G. et al. Smoldering myocarditis following immune checkpoint Blockade. J. Immunother Cancer. 5, 91 (2017).
Article Google Scholar
Champion, S. N. & Stone, J. R. Immune checkpoint inhibitor associated myocarditis occurs in both high-grade and low-grade forms. Mod. Pathol. 33, 99–108 (2020).
Article CAS Google Scholar
Nishimura, H. et al. Autoimmune dilated cardiomyopathy in PD-1 receptor-deficient mice. Science 291, 319–322 (2001).
Article ADS CAS Google Scholar
Gergely, T. G. et al. Characterization of immune checkpoint inhibitor-induced cardiotoxicity reveals interleukin-17A as a driver of cardiac dysfunction after anti-PD-1 treatment. Br. J. Pharmacol. 180, 740–761 (2023).
Article CAS Google Scholar
Xiong, X., Yu, M., Wang, D., Wang, Y. & Cheng, L. Th17/Treg balance is regulated by myeloid-derived suppressor cells in experimental autoimmune myocarditis. Immun. Inflamm. Dis. 11, e872 (2023).
Article CAS Google Scholar
Bednar, K. J., Lee, J. H. & Ort, T. Tregs in autoimmunity: insights into intrinsic brake mechanism driving pathogenesis and immune homeostasis. Front. Immunol. 13, 932485 (2022).
Article CAS Google Scholar
Fontenot, J. D., Gavin, M. A. & Rudensky, A. Y. Foxp3 programs the development and function of CD4 + CD25 + regulatory T cells. Nat. Immunol. 4, 330–336 (2003).
Article CAS Google Scholar
Nakamura, K., Kitani, A. & Strober, W. Cell contact-dependent immunosuppression by CD4(+)CD25(+) regulatory T cells is mediated by cell surface-bound transforming growth factor beta. J. Exp. Med. 194, 629–644 (2001).
Article CAS Google Scholar
Green, E. A., Gorelik, L., McGregor, C. M., Tran, E. H. & Flavell, R. A. CD4 + CD25 + T regulatory cells control anti-islet CD8 + T cells through TGF-beta-TGF-beta receptor interactions in type 1 diabetes. Proc. Natl. Acad. Sci. U S A. 100, 10878–10883 (2003).
Article ADS CAS Google Scholar
Fahlén, L. et al. T cells that cannot respond to TGF-beta escape control by CD4(+)CD25(+) regulatory T cells. J. Exp. Med. 201, 737–746 (2005).
Article Google Scholar
Kim, H. J., Hwang, S. J., Kim, B. K., Jung, K. C. & Chung, D. H. NKT cells play critical roles in the induction of oral tolerance by inducing regulatory T cells producing IL-10 and transforming growth factor beta, and by clonally deleting antigen-specific T cells. Immunology 118, 101–111 (2006).
Article CAS Google Scholar
Mokkala, K. et al. Gut microbiota richness and composition and dietary intake of overweight pregnant women are related to serum Zonulin concentration, a marker for intestinal permeability. J. Nutr. 146, 1694–1700 (2016).
Article CAS Google Scholar
Chelakkot, C., Ghim, J. & Ryu, S. H. Mechanisms regulating intestinal barrier integrity and its pathological implications. Exp. Mol. Med. 50, 1–9 (2018).
Article CAS Google Scholar
Gil-Cruz, C. et al. Microbiota-derived peptide mimics drive lethal inflammatory cardiomyopathy. Science 366, 881–886 (2019).
Article ADS CAS Google Scholar
Gergely, T. G., Drobni, Z. D., Sayour, N. V., Ferdinandy, P. & Varga, Z. V. Molecular fingerprints of cardiovascular toxicities of immune checkpoint inhibitors. Basic. Res. Cardiol. 120, 187–205 (2025).
Article CAS Google Scholar
Bartolomaeus, H. et al. Short-Chain fatty acid propionate protects from hypertensive cardiovascular damage. Circulation 139, 1407–1421 (2019).
Article CAS Google Scholar
Li, X. et al. Trimethylamine N-Oxide exacerbates cardiac fibrosis via activating the NLRP3 inflammasome. Front. Physiol. 10, 866 (2019).
Article Google Scholar
Huang, K. et al. 2-O-β-d-Glucopyranosyl-l-ascorbic acid, an ascorbic acid derivative isolated from the fruits of Lycium Barbarum L., modulates gut microbiota and palliates colitis in dextran sodium Sulfate-Induced colitis in mice. J. Agric. Food Chem. 67, 11408–11419 (2019).
Article CAS Google Scholar
He, P. et al. The Maca protein ameliorates DSS-induced colitis in mice by modulating the gut microbiota and production of SCFAs. Food Funct. 14, 10329–10346 (2023).
Article CAS Google Scholar
Zuo, Z. et al. Investigation of gut microbiota disorders in Sepsis and Sepsis complicated with acute Gastrointestinal injury based on 16S rRNA genes illumina sequencing. Infect. Drug Resist. 16, 7389–7403 (2023).
Article CAS Google Scholar
Henke, M. T. et al. Ruminococcus gnavus, a member of the human gut Microbiome associated with crohn’s disease, produces an inflammatory polysaccharide. Proc. Natl. Acad. Sci. U S A. 116, 12672–12677 (2019).
Article ADS CAS Google Scholar
Tandon, R. et al. Revisiting the pathogenesis of rheumatic fever and carditis. Nat. Rev. Cardiol. 10, 171–177 (2013).
Article CAS Google Scholar
Schmutzler, L., Mirna, M., Hoppe, U. C. & Lichtenauer, M. From Streptococcal pharyngitis/tonsillitis to myocarditis: A systematic review. J Cardiovasc. Dev. Dis 9, 170 (2022).
Cuccherini, B. et al. Bacteremia and skin/bone infections in two patients with X-linked agammaglobulinemia caused by an unusual organism related to flexispira/helicobacter species. Clin. Immunol. 97, 121–129 (2000).
Article CAS Google Scholar
Freeman, A. F. & Holland, S. M. Persistent bacterial infections and primary immune disorders. Curr. Opin. Microbiol. 10, 70–75 (2007).
Article CAS Google Scholar
Rogier, R. et al. Aberrant intestinal microbiota due to IL-1 receptor antagonist deficiency promotes IL-17- and TLR4-dependent arthritis. Microbiome 5, 63 (2017).
Article Google Scholar
Liu, H. et al. IBD-Associated Atg16L1T300A polymorphism regulates commensal microbiota of the intestine. Front. Immunol. 12, 772189 (2021).
Article CAS Google Scholar
Liu, B. et al. Two-Sample Mendelian randomization analysis investigates causal associations between gut microbial genera and inflammatory bowel disease, and specificity causal associations in ulcerative colitis or crohn’s disease. Front. Immunol. 13, 921546 (2022).
Article CAS Google Scholar
DeSantis, T. Z. et al. Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl. Environ. Microbiol. 72, 5069–5072 (2006).
Article ADS CAS Google Scholar

Download references

Funding

This study was supported by Health Commission of Shanxi Province [grant number 2020071].

Author information

Huili Cao and Huwei Dai contributed equally.

Authors and Affiliations

Department of Cardiovascular, The Second Hospital of Shanxi Medical University, Taiyuan, Shanxi, China
Huili Cao, Yajing Zhao & Bin Yang
Shanxi Medical University, Taiyuan, Shanxi, China
Huwei Dai, Songshan Li, Zeeshan Afzal, Xinyan Wang, Zeyu Wen & Kaiyong Xiao
Department of Endocrinology and Metabolism, The Second Hospital of Shanxi Medical University, Taiyuan, Shanxi, China
Jin Li

Authors

Huili Cao
View author publications
Search author on:PubMed Google Scholar
Huwei Dai
View author publications
Search author on:PubMed Google Scholar
Songshan Li
View author publications
Search author on:PubMed Google Scholar
Zeeshan Afzal
View author publications
Search author on:PubMed Google Scholar
Xinyan Wang
View author publications
Search author on:PubMed Google Scholar
Zeyu Wen
View author publications
Search author on:PubMed Google Scholar
Kaiyong Xiao
View author publications
Search author on:PubMed Google Scholar
Yajing Zhao
View author publications
Search author on:PubMed Google Scholar
Jin Li
View author publications
Search author on:PubMed Google Scholar
Bin Yang
View author publications
Search author on:PubMed Google Scholar

Contributions

J.L. and B.Y. conceived and designed the experiments. S.L., K.X. and Z.W. performed the experiments. X.W. and Y.Z. analyzed the data. H.C., H.D. and Z.A. wrote the paper. All authors contributed to the article and approved the submitted version.

Corresponding authors

Correspondence to Jin Li or Bin Yang.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Cao, H., Dai, H., Li, S. et al. Abnormal gut microbiota may cause PD-1 inhibitor-related cardiotoxicity via suppressing regulatory T cells. Sci Rep 15, 20547 (2025). https://doi.org/10.1038/s41598-025-05635-4

Download citation

Received: 14 August 2024
Accepted: 03 June 2025
Published: 01 July 2025
DOI: https://doi.org/10.1038/s41598-025-05635-4